二尖瓣置换影响左心室流场的体外实验研究
IN VITRO EXPERIMENTAL STUDY ON THE LEFT VENTRICULAR FLOW OF MITRAL VALVE REPLACEMENT
-
摘要: 二尖瓣置换术是通过植入人工机械瓣或生物瓣替换病变原生瓣膜的重要外科治疗手段. 本研究通过体外实验平台, 系统探究了生物瓣与机械瓣置换对左心室血流动力学的影响, 重点关注涡流结构与压力分布的差异. 实验采用基于成人心脏CT数据重建的3D打印的硅胶左心室模型, 结合完整保留的动物原生二尖瓣、瓣下结构(包括腱索与乳头肌)和主动脉, 构建了近生理条件的体外循环模拟系统. 通过三维层析粒子图像测速技术(Tomo PIV)对4种工况(原生瓣、生物瓣和相对于原生瓣膜平行/垂直放置的机械瓣)的流场进行高分辨率三维动态测量, 并结合基于物理信息神经网络(PINN)的曲面动边界压力场重构技术, 定量分析了不同瓣膜对左心室血流动力学参数的影响. 实验结果显示, 生物瓣在舒张期形成与原生瓣方向一致的完整涡环结构, 但其射流速度显著高于原生瓣(0.93 vs. 0.62 m/s), 舒张末期心室壁面仍维持较高的压力梯度, 可能增加流体对心室壁的作用力. 机械瓣虽保持与原生瓣相近的跨瓣压差, 但其双叶设计引发3股射流相互作用, 导致涡流方向逆转且流场复杂度显著增加, 尤其在垂直放置时涡量分布呈现明显不对称性. 此外, 两类人工瓣膜均未能使涡结构充分延伸拓展至心尖区域, 可能增加血液滞留与血栓风险. 原生瓣膜则表现出最优的血流效率, 流体在舒张期充分冲刷心尖区域, 减少了血流在心尖处停滞时间. 本研究结合了Tomo PIV与PINN技术, 突破了传统压力场重构方法在曲面动边界条件下的局限性, 实现了人工瓣膜置换后左心室三维流场与压力场的同步动态解析. 研究结果为临床瓣膜选型提供了血流动力学依据: 与机械瓣相比, 生物瓣更有利于维持有序流场, 但需关注其高剪切应力对血细胞的潜在损伤; 机械瓣虽耐久性更佳, 但复杂流场可能加剧心室不良重塑. 未来研究将进一步纳入个体化解剖变异与病理参数, 以优化瓣膜设计并提升长期手术预后.Abstract: Mitral valve replacement (MVR) is a critical surgical intervention for severe valvular dysfunction, yet the hemodynamic consequences of prosthetic valve selection remain incompletely understood. This study investigates the impact of bioprosthetic and bileaflet mechanical valves on left ventricular (LV) flow patterns and pressure dynamics using an in vitro platform replicating physiological conditions. A 3D-printed silicone LV model, integrated with porcine mitral valve (including chordae tendineae and papillary muscles) and aorta, was cyclically actuated via a mock circulatory loop. Four configurations were tested: native valve, bioprosthetic valve, and mechanical valves oriented parallel/vertical to the native annulus. Time-resolved 3D flow fields were captured using tomographic particle image velocimetry (Tomo PIV), while a physics-informed neural network (PINN) framework was employed to reconstruct dynamic pressure fields under moving boundary conditions. Key findings revealed distinct hemodynamic profiles: the bioprosthetic valve generated a coherent diastolic vortex ring aligned with native flow direction, but exhibited elevated transvalvular jet velocity (0.93 m/s vs. 0.62 m/s) . This may impose higher fluid-induced wall stress, potentially triggering adverse remodeling. In contrast, the mechanical valve exhibited diastolic pressure distribution comparable to the native state, but flow reversal occurred due to interactions among triple jets generated by its bileaflet structure, particularly in the vertical orientation where asymmetric vorticity distribution dominated. Both prosthetic types failed to propagate vortices to the apical region, increasing stasis risk. The native valve demonstrated optimal efficiency, with apical vortex penetration minimizing blood residence time. By integrating Tomo PIV with PINN-based pressure reconstruction, this study overcomes limitations of traditional methods in handling deforming boundaries, enabling synchronized 3D flow-pressure analysis post-MVR. Clinically, compared with mechanical valves, bioprosthetic valves better maintain organized flow patterns but require careful monitoring of high shear stress-induced potential damage to blood cells; mechanical valves, while exhibiting superior durability, may exacerbate adverse ventricular remodeling due to complex flow patterns. These insights underscore the need for personalized valve selection based on hemodynamic profiling. Future work will incorporate patient-specific geometries and pathological parameters to refine predictive models and prosthesis design.